Inaccuracies in the Tabulated Input and Output in Intensive Care: A Potential Risk to Optimal Patient Care

Background:

Fluid management of patients can be challenging in the best circumstances. The relationship of total fluid input to total output is commonly used to guide therapeutic decisions involving the use of diuretics and/or fluids. Careful management of fluid balance is a cornerstone of management of patients with critical illness including acute respiratory failure, acute renal failure, and congestive heart failure.

We noted a number of incidents in the ICU in which daily total values for input and output tabulated on the ICU flow sheet appeared inconsistent with expectations based on clinical observations. On review of the details on the flow sheet, we determined that large mathematical errors had occurred, at times representing more than 1 L of discrepancy. As this degree of error in the tabulation of fluid balance could well lead to suboptimal therapeutic decisions, we undertook a study to determine the prevalence and significance of this problem.

Methods:

Every flow sheet in an 18‐bed academic tertiary‐care ICU was reviewed for a full week. All the data points from the flow sheets were entered into a spreadsheet where hourly and daily input and output calculations were done. The calculated totals were then compared to the totals entered on the flow sheets. Analyses of both individual errors and 24‐hour total variances were performed separately for input and for output. Variances were then carefully reviewed to attempt to understand where mistakes occurred and to determine whether a pattern existed.

Results:

One hundred and twenty‐six flow sheets were reviewed, of which 71% had an error in the addition of fluid input or output. Four percent had an error of greater than 1 L. Of those, 60% were related to miscalculations made at shift change.

Conclusions:

It is not uncommon for flow sheets in the ICU to have large mathematical miscalculations that lead to erroneous daily totals that are potentially clinically significant. These errors tend to occur during calculation of totals at shift changes.

The clinical impact of such errors is yet to be determined but could easily lead to suboptimal dosing of diuretic or fluid therapy.

Extrapolating a 4% frequency of errors of 1 L or more, it would be anticipated that such an error would occur in 1 patient per day in an our 18‐bed ICU.

It is of note that only simple errors in the addition of fluid input or output were captured by our study. Errors in which fluid input or output is not recorded at all (error of omission) or in which an incorrect hourly subtotal is recorded (data entry error) may also be prevalent.

Author Disclosure:

K. S. Fleming, None; C. A. Block, None.